Customer Orientation

Technology and Customer Orientation

Graphics courtesy of Marketing-91Opens in new window

For effective and efficient orientation toward the customer, thorough market and customer knowledge is a prerequisite. Only then is it organizationally possible to realize an individualized offering model and effective customer care.

If this information is collected and analyzed in appropriate ways, a potential of enormous worth arises. Often the information that already exists in the company simply needs to be evaluated.

‘When diverse information about the customers exist in various departments, there’s nothing better to do than to evaluate it from various perspectives’, says Johannes Wechsler, Managing Director of the software provider ApplixOpens in new window.

A similar point of view is held by Arthur Parker, Vice President of Global Business Intelligence Solutions EMEA at IBM, ‘Customer relationships should be systematically developed, rather than arising out of a chain of coincidences. From analyzing the information on hand, future sales are more readily foreseeable’ (Gobbel, 2001, p. 27)

With the engagement of key technologies like data warehousing and data mining this information can be evaluated and administrated in a focused manner.

1.     Data Warehouse

Augmented, a data warehouse can be compared to reporting and analyses tools. Information is well sorted and quickly accessible in storage (www.bbdo.deOpens in new window). Connections and possible relationships can be established very rapidly. As an example in practice, many large financial institutions have compiled data warehouses for their commercial customer relationships.

Financial statements can be incorporated into the data warehouse when they are received annually, which provides the lending officers with quality information in order to assess future borrowing needs for their customers. This information, when combined with demographic data and deposit information, can serve as a useful sourcing tool for new business opportunities. The key is to be able to successfully mine the data for future use.

2.     Data Mining

It is used primarily in two cases. In the first case, data mining can be used via intelligent statistical procedures where adequate segmentations can be identified, which enable the output of a targeted item of information.

In the second case, data mining can be used for the calculation of purchase probabilities. By means of statistical procedures and neural networks it is possible, for example, on the basis of specific parameters, to show the repeat purchase probability of customers or potential customers (www.bbdo.deOpens in new window).

The three most important procedures of data mining for the forecasting and describing of data are segmentation, classifying, and associating (Schinzer, Bange & Mertens, 1999, p. 104 – 107).

Each of these procedures in turn makes use of various techniques. The method used is determined by the searched for pattern, and heavily exhibits therefore a problem orientation that is situationally conditioned.

The more detailed description of methods that follows, therefore, limits itself to the most commonly used techniques. These are neural networks, cluster analyses, and decision tree procedures for finding sequence patterns and association rules.

3.     Segmentation

As far as applications in business are concerned, segmentation has notable advantages. From out of huge data stocks a company can filter out a potential customer.

In the context of a target group, marketing a product offer can be specialized and the precision of demand predictions can be enhanced, which are necessary for production planning and the calculation of stock capacity for new products. For segmentation, statistical cluster analyses procedures and a special form of neural networks, so called kohonen nets, are mainly used.

4.    Classification and Regression

The task of classification procedures is to assign elements, whose characteristics are unknown, to existing classes, or to make predictions concerning unknown values of an existing data stock.

On the basis of shared attributes of the elements in each class, a model to describe the classes is developed, which during the course of the classification ends up being portrayed in structures or rules.

Classification procedures are used in database marketing, for example, to increase efficiency of direct advertising, to improve product range design, and to prevent loss of customers. The discovery of credit card fraud and the evaluation of default risk are further areas in which data mining is useful.

A classification-related procedure of regression is often used for the creation of prediction models. An object is here classified on the basis of its attribute, though not its class, rather a relevant index like sales turnover or profitability, for example, demand functions for a product in supermarkets, which involve many variables (like location in the business, the amount of shelving space, price, marketing measures, customer structure of the business, time of year, and time of day).

5.     Associating

With the help of associating, dependencies between different data records can be filtered out. Correlations between concurrently occurring items are described. A typical use of association is via shopping basket analysis.

By viewing a great number of transactions, purchasing trends should be apparent, which will reveal the purchasing behavior of customers. Firms that successfully use the associating technique will be in a better position to determine the stage of the life cycle of their customers.

Products offers, product positioning, or targeted marketing, can all be improved using this method. It should be mentioned here that this is a primary mechanism for improving the net profit margin of a firm. There are countless examples of poor marketing segmentation efforts leading to the wasteful production of marketing material that is sent to uninterested potential customers.

The better a firm is at identifying potential customers for a marketing campaign, the better the company will do in terms of achieved revenue gains as well as the avoidance of unnecessary costs. Companies as varied as the grocery chain Tesco and the online retailer Amazon.com have used the associating technique successfully over the last few years.